Effects of deforestation and reforestation on landscape spatial structure in boreal Saskatchewan, Canada

Abstract

Monitoring landscape spatial structure has been recommended to detect degenerative trends in forest conditions, but such data have not been reported for boreal regions of North America under the influence of agricultural expansion. In this study, I analyzed the spatial characteristics of wooded area for several boreal landscapes in western Canada that exhibited persistent deforestation over approximately three decades prior to 1990. Patch area and perimeter metrics were quantified from topographic map chronosequences for the Halkett Lake, Mistatim, Debden and Bjorkdale areas of central Saskatchewan. The four landscapes varied from approximately 75–35% wooded and spanned a descending gradient in the proportion of the total area protected from land use change by federal or provincial legislation. Largest patch size and degree of coherence exhibited consistent temporal declines as deforestation proceeded within each landscape. When comparisons were made across the four landscapes within fixed time periods, largest patch size index and effective mesh size decreased as the proportion of land area wooded declined. A linear relationship with slope greater than unity was evident between the largest patch size index and the proportion of land area wooded for 1975/1976 ( r 2=0.99) and 1990 ( r 2=0.99). Metrics based on the size of the largest patch or the sum of squared patch sizes proved useful as baseline measures of landscape spatial structure. Number of patches, perimeter-to-area ratio, mean patch size and related metrics were less valuable as they exhibited variable responses with progressive deforestation. The temporal and spatial analyses yielded evidence that fragmentation exacerbated the effects of absolute wooded area losses. Deforestation disproportionately reduced the sizes of the largest patches, a finding with important management implications. Efforts to reverse deforestation within these landscapes should place priority on expanding large patches. Reforestation of agricultural lands has been proposed to offset carbon emissions from fossil fuel burning. Simulated reforestation with dispersed small patches resulted in little change in fragmentation as measured with landscape structure metrics. Simulated reforestation with larger blocks adjacent to existing large patches initiated reversal of the historic trends for largest patch size and degree of coherence. Increasing the size of provincial forests or protected areas, or restoring forest cover adjacent to them, would serve to expand the size of large forest patches. This reforestation strategy could provide supplementary benefits beyond carbon sequestration by reversing the degenerative process of fragmentation that has impaired forest wildlife and ecosystem processes.